CN105069682A - Method for realizing mass sensitivity-based incentive mechanisms in mobile crowdsourcing systems - Google Patents

Abstract

The invention discloses a method for realizing mass sensitivity-based incentive mechanisms in mobile crowdsourcing systems. The method comprises the following steps: distributing a task by a distributor through a platform; preliminarily screening participants by the platform; participating by the screened participants through a competitive bidding way; sending data to the platform by the participants which satisfy secondary screening of the platform; carrying out quality detection on the data by the platform so as to obtain data reliability; scoring the data by the distributor; and finally closing the deal. According to the method provided by the invention, a game mechanism-based validation method is adopted and the scores of the distributor are validated through players. Finally, the platform updates the reputation degrees of the distributor and the players through the factors such as the reliability value of the data, the scores of the distributor, the credit rank of the distributor and the like. According to the method provided by the invention, the behaviors of the users in the mobile crowdsourcing systems can be effectively stimulated and the data quality is improved.

Description

Based on the incentive mechanism implementation method of mass-sensitive in a kind of mobile mass-rent system

Technical field

The present invention relates to the incentive mechanism implementation method based on mass-sensitive in a kind of mobile mass-rent system, belong to data mining technology field.

Background technology

Along with the development of modern smart mobile phone, the embedded type sensor equipment of mobile phone is more and more cheap, and more and more abundanter, function is also more powerful, therefore, provides a good mode and carrys out image data, and also create the application of many mobile mass-rent perception.In mobile mass-rent system, an important problem is incentive mechanism.

But, how by incentive mechanism, satisfied one side can make participant provide a large amount of and data reliably, ensures that the carrying out of publisher's honesty feeds back on the other hand, also to ensure the profit of platform simultaneously, become the hot topic research topic in current mobile mass-rent system.

Incentive mechanism in traditional mobile mass-rent system is mainly divided three classes, that is: the first kind is the mode by amusement, as data acquisition is carried out in the game carried out based on geographic position; Equations of The Second Kind is the mode by peering service, and namely publisher also plays the role of participant simultaneously; 3rd class is then the mechanism based on reverse auction, and participant is realized by competitive bidding.But method in the past only only considered enthusiasm and the reliability of participant, and often platform and publisher are integrated, and the scope therefore involved by this kind of application compares limitation.And the present invention can solve problem above well.

Summary of the invention

The object of the invention there are provided the incentive mechanism implementation method based on mass-sensitive in a kind of mobile mass-rent system, this mechanism effectively combines the mechanism such as reverse auction, Game Authentication, the renewal of popularity degree, first the selection of both parties is realized by reverse auction mechanism, by the method for Game Authentication, whether publisher is maliciously passed judgment on, finally by the calculating of the aspect such as credit grade of the confidence level of the data to participant, the scoring of publisher and publisher, platform upgrades the popularity degree of both parties.Compared with current mechanism, the present invention mainly comprises: first, utilize the advantage of reverse auction, make the interests foundation contact of publisher, platform, participant three, relative to traditional mechanism, invention increases the Filtering system of platform for participant and the computing mechanism of data reliability, effectively under the condition meeting participant's requirement, ensure that the interests of platform and publisher.Secondly, the present invention has taken into full account the scoring of publisher, introduce Game Authentication mechanism and avoid the malice of publisher scoring behavior (publisher can be marked by malice improve the income of oneself), compared by the scoring of participant in game and the scoring of publisher, thus reflect actual scoring really.The present invention introduces the counting system of popularity degree simultaneously, the data quality problem only considering participant side is calculated relative to popularity degree in conventional driver mechanism, in this mechanism, the counting system of popularity degree divide into participant's popularity degree and publisher's popularity degree two class, wherein participant's popularity degree upgrades and contains the confidence level of data and the scoring of publisher, publisher's popularity degree upgrades the credit grade containing publisher, by concrete calculating, make popularity degree effectively can reflect the actual conditions of participant and publisher.Popularity degree all can have an impact in each step of mechanism, and popularity degree is higher, and the benefit of user is also higher, therefore, it is possible to the behavior of effective excitation participant and publisher.

The technical solution adopted for the present invention to solve the technical problems is: the invention provides in a kind of mobile mass-rent system based on participant, platform, the incentive mechanism based on mass-sensitive of publisher's three aspect factor, reverse auction mechanism is dissolved in platform by the method, consider the revenue relations of three simultaneously, and add the malice scoring behavior that Game Authentication mechanism avoids publisher, finally according to the confidence level of data, the scoring of publisher and the credit grade of publisher, the popularity degree of participant and publisher is upgraded, popularity degree each step in this mechanism all can have an impact, participant and publisher is encouraged to obtain higher income by improving popularity degree with this.

Method flow:

Step 1: reverse auction;

Step 1-1: the task division of oneself is little task that is simple, that be easy to mobile terminal realization by publisher, is published on platform, and pays certain amount of money P _puto platform;

Step 1-2: publisher has certain restriction requirement to task, this platform needs to carry out a preliminary screening to participant; Including but not limited to region screening, the age screens, and sex is screened, industry screening etc.;

Step 1-3: carry out the participant after preliminary screening provides oneself bidding price b according to actual conditions _i;

Step 1-4: platform is according to rank price S _icarry out postsearch screening;

Step 1-5: platform notice competitive bidding successful m name participant, m name participant sends to platform by after information;

Step 1-6: the confidence level of platform to data calculates;

Step 1-7: publisher receives the data that user sends, carries out evaluation e to data _pu.If e _pufor difference is commented, then enter step 2, otherwise enter step 2-6;

Step 2: Game Authentication;

Step 2-1: platform carries out the authentication method based on game;

Step 2-2: n name player recruited by platform, is c to the amount of money award of every player;

Step 2-3: player carries out scoring to data and draws e _g;

Step 2-4: platform is by e _gwith e _puvalue compares, and judges whether publisher is malice scoring;

Step 2-5: after thinking after platform is passed judgment on that publisher is correct scoring, platform will return the certain amount of money of publisher, and can not pay participant's amount of money; After thinking after platform is passed judgment on that publisher is malice scoring, then publisher can produce extra expense;

Step 2-6: platform is by amount of money b _ibe given to participant m _i;

Step 3: popularity degree upgrades;

Step 3-1: platform is according to the scoring of publisher, and the confidence level of data, the credit grade of publisher upgrades the popularity degree of both sides.

Step 1-1 of the present invention can be divided into two steps, comprising: the first step, and the task division of oneself is the little task being simply easy to mobile terminal realization by publisher, is published on platform.Second step, publisher's payment platform amount of money, wherein publisher payment P _pucalculating comprise: the bidding price of participant, the profit of platform, carry out the cost of authenticate of playing.P _pucomputing formula be: wherein, b _ifor m name participant bidding price separately, if there is the transaction of failure, then corresponding b _ivalue strain be 0.P _platfor the profit amount of money of platform, G is the expense that platform carries out the mechanism of playing.Platform profit amount of money P _platcomputing formula be:

Wherein $per=\left\{\begin{array}{cc}0& {\mathrm{RE}}_{pu}\&GreaterEqual;{\mathrm{threRE}}_{pu}\\ (1-{\mathrm{RE}}_{pu})\&times;k& {\mathrm{RE}}_{pu}<{\mathrm{threRE}}_{pu}\end{array}\right.,$ RE _pufor the popularity degree of publisher, threRE _pufor the threshold value of platform defined, k is constant, and k gets 5% in the present invention.The computing formula of the expense G of game mechanism is: G=n × c (and meet: G≤0.5b _i), wherein n is that platform carries out the players number of recruiting of playing, and c is the award that platform gives player.

Comprise rank price S in step 1-4 of the present invention _icalculating, rank price S _icomputing formula be: S _i=b _i× (1-RE _pa), wherein, RE _pafor the popularity degree of participant, RE _pabe worth higher, rank price is just relatively lower.Rank price S _ionly for rank, participant's real income is still b _i.Platform is according to rank price S _irank, selects m name participant in order successively, and meets

Platform notice competitive bidding successful m name participant in step 1-5 of the present invention, m name participant sends to platform by after information.As long as transmission data, platform will give the virtual award of participant (that is: badge etc.).

Comprise the calculating to data reliability in step 1-6 of the present invention, comprise two aspects, that is: 1) the quality coefficient Q of data; 2) the confidence level T of participant.To the computing formula of the confidence level Tod of data be: Tod=a × Q+ (1-a) × T, Tod belongs to 0 ~ 1, wherein a is constant, and Q is the quality coefficient of data, and T is the confidence level of participant.In the present invention, because weight is higher shared by the quality of data, so a gets 0.7.The quality coefficient Q of data is calculated by existing video image quality detection technique, sound quality detection technique etc.Wherein the value of Q belongs to 0 ~ 1.The computing formula of confidence level T is: wherein TE is the set of the speciality of required by task, and PE is the set of the speciality that participant has, and n gathers the number comprised.

The scoring e of publisher to data is comprised in step 1-7 of the present invention _pu, its standards of grading are:

If publisher is evaluated as non-difference when commenting, then platform is by amount of money b _ibe given to participant m _i, platform upgrades the popularity degree of both sides.If publisher is evaluated as difference when commenting, then Game Authentication mechanism will be carried out.

In step 2-1 of the present invention, platform carries out based on the authentication method of game is that platform is marked to data by recruiting player, give the certain amount of money of player to reward, the scoring of player and the scoring of publisher are compared, determines whether publisher has the behavior of malice evaluation.

In step 2-2 of the present invention, n name player recruited by platform, is c (that is: c is constant) to the amount of money award of every player.The wherein selection demand fulfillment G=nc≤0.5b of n _i, to alleviate the extra amount burden that participant bears.

E in step 2-3 of the present invention _gcomputing formula be: wherein e _gvalue belong to 0 ~ 1, e _girepresent n-th _iplayer is to the scoring of data for name, and belongs to 0 ~ 1.

E in step 2-4 of the present invention _gwith e _puthe comparison of value, judges that whether publisher is the comparative approach of malice scoring, and to be the actual scoring of calculating participant popularity when spending be e:

A.e _g∈ (0,0.6], e _pu∈ (0,0.6] time, when namely player and publisher are poor commenting, illustrate that publisher is correct scoring.

As Δ=e _g-e _pu>=0.3, get e=(e _g+ e _pu)/2,

As Δ=e _g-e _pu<0.3, gets e=e _pu

B.e _g∈ (0.6,0.8], e _pu∈ (0,0.6] time, when namely player's scoring is commented in being

As Δ=e _g-e _pu>=0.5, get e=e _gillustrate that publisher is malice scoring

As Δ=e _g-e _pu∈ [0.2,0.5), get e=(e _g+ e _pu)/2, now as e=(e _g+ e _pu)/2 ∈ (0.6,0.8] time, illustrate that publisher is malice scoring; As e=(e _g+ e _pu)/2 ∈ (0,0.6] time, illustrate that publisher is correct scoring.

As Δ=e _g-e _pu∈ (0,0.2), gets e=e _pu, illustrate that publisher is correct scoring.

C.e _g∈ (0.8,1.0], e _pu∈ (0,0.6] time, namely player is favorable comment, when publisher is for poor commenting,

Get e=e _g, illustrate that publisher is malice scoring.

In step 2-5 of the present invention after thinking after platform is passed judgment on that publisher is correct scoring, platform will return publisher b _ithe amount of money of+n × c, wherein b _ithe amount of money be originally will to participant m _ithe amount of money, n × c carries out the expense (if namely user correctly marks, just do not need for game machine makes expense) of Game Authentication mechanism as platform; After thinking after platform is passed judgment on that publisher is malice scoring, then the expense of publisher is b _i+ 2n × c, wherein b _ifor paying participant m _i, n × c pays player, and another n × c is as the extra punishment of platform for publisher.

To participant's popularity degree RE in step 3-1 of the present invention _pacalculating comprise the data reliability Tod of participant and the scoring of publisher, and RE _pabelong to 0 ~ 1, its computing formula for for:

(1) as the evaluation e that publisher provides _pufor favorable comment or in comment time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein threTod is the judge threshold value of data reliability, gets threTod=0.5, k ₁and k ₂be the constant that determined by platform and all belong to 0 ~ 1.

(2) as the evaluation e that publisher provides _puduring for poor commenting,

A. when evaluation result for publisher for correctly evaluate time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein e is that platform is to e _gand e _pucompare the result of rear generation.

B. when evaluation result is evaluated for malice for publisher,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2}{1+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2}.$

To the popularity degree RE of publisher _pucalculating comprise the credit grade of publisher, RE _pubelong to 0 ~ 1, its computing formula is:

First the current credit grade of this subtask of publisher is defined (wherein non-malicious difference comments number to comprise the favorable comment number of publisher, in when commenting number and publisher's difference to comment but platform is judged as the number of correctly marking), x ∈ [-1,1],

The current popularity degree of publisher is calculated according to the credit grade of publisher b, c are the constant that platform specifies.Then the RE after renewal is calculated with sliding window algorithm _puvalue, then the popularity degree RE of publisher after upgrading _pu=α RE _pu+ (1-α) cRE _pu, α is the constant that platform is formulated.

Beneficial effect:

1, the present invention is in conjunction with the advantage of reverse auction mechanism, effectively can guarantee the interest relations between participant, platform, publisher three, introduce twice screening process of platform simultaneously, effectively can ensure the confidence level of participant's data, making on the one hand can while ensureing that participant obtains income, make platform and publisher spend less expense as much as possible, obtain high-quality data.

2, the present invention is in order to avoid the malice scoring behavior of publisher, and take the method for Game Authentication, platform is evaluated by employing the scoring of player to publisher, by concrete judgment criteria, comprises Δ=e _g-e _puinterval division standard, effectively ensure that the validity that publisher marks.

3, this invention takes the calculating more new system of popularity degree, by the scoring to publisher, participant sends the calculating of the confidence level of data and the credit grade of publisher, point different situations complete and upgrade the popularity degree of participant and publisher, effectively by popularity degree, they are connected, and popularity degree plays important role in this incentive mechanism, effectively incentive action can be produced to the behavior of both sides.

4, three kinds of traditional machine-processed advantages have been merged in the present invention, consider the motivator of participant, platform, publisher three aspect simultaneously, and add the more new system of Game Authentication mechanism and user's popularity degree, while the guarantee quality of data, also can ensure the profit of platform and the enthusiasm of publisher, effectively expand the range of application of incentive mechanism.

Accompanying drawing explanation

Fig. 1 is the structural representation of present system.

The schematic flow sheet of Fig. 2 for using Game Authentication mechanism (gamification) feedback to publisher to pass judgment in invention.

Fig. 3 is the schematic flow sheet upgrading participant (participant) popularity degree in the present invention.

Fig. 4 is the schematic flow sheet upgrading publisher (publisher) popularity degree in the present invention.

Embodiment

Below in conjunction with Figure of description, the invention is described in further detail.

Symbol of the present invention and implication thereof comprise:

P pu	Publisher needs the total charge paid	RE pu	The popularity degree of publisher
				P plat	The platform profit amount of money	threRE pu	The popularity degree threshold value of publisher
b i	Participant m iBidding price	Tod	Data reliability
				G	Platform carries out the expense of the mechanism of playing	Q	Quality of data coefficient
m	The number of participant	T	The confidence level of participant
				n	The number of player	e pu	Publisher is to the scoring of data
c	The amount of money of player paid by platform	e g	Player is to the scoring of data
				S i	Rank price	e	e pu，e gActual scoring relatively
RE pa	The popularity degree of participant	cRE pu	The current popularity degree of publisher
				x	The credit grade of publisher

As shown in Figure 1, total system of the present invention is mainly divided into three parts: Part I and reverse auction part, exchange for the interests completed between participant, platform, publisher three, and add twice screening process of platform, under making to meet the condition that participant requires, ensure the interests of platform and publisher; Part II is Game Authentication mechanism, for verifying that whether publisher is malice scoring, by the judgment criteria formulated, makes it possible to draw a realistic scoring according to the scoring of player and publisher; Part III is the renewal of user's popularity degree, is divided into the popularity degree of participant and publisher, contains the calculating of scoring to data reliability, publisher and publisher's credit grade.The operation workflow of system is specific as follows: step 1: the task division of oneself is the little task being simply easy to mobile terminal realization by publisher, is published on platform, and pays certain amount of money P _puto platform.Wherein wherein b _ifor m name participant bidding price separately, P _platfor the profit amount of money of platform, G is the expense that platform carries out the mechanism of playing.

$P_{plat}=\underset{i=1}{\overset{m}{\&Sigma;}}{b}_i\&times;per,$ Wherein $per=\left\{\begin{array}{cc}0& {\mathrm{RE}}_{pu}\&GreaterEqual;{\mathrm{threRE}}_{pu}\\ (1-{\mathrm{RE}}_{pu})\&times;k& {\mathrm{RE}}_{pu}<{\mathrm{threRE}}_{pu}\end{array}\right.$

Wherein RE _pufor the popularity degree of publisher, threRE _pufor the threshold value of platform defined, k is constant, if there is the transaction of failure, then and corresponding b _ivalue strain be 0.

G=n × c (G≤0.5b _i), wherein n is that platform carries out the players number of recruiting of playing, and c is the award that platform gives player.

Step 2: publisher has certain restriction requirement to task, therefore platform needs to carry out a preliminary screening to participant, and including but not limited to region screening, the age screens, and sex is screened, industry screening etc.

Step 3: carry out the user after preliminary screening provides oneself bidding price b according to actual conditions _i.

Step 4: platform is according to S _icarry out postsearch screening.The rank price S of platform _i=b _i× (1-RE _pa), wherein, RE _pafor the popularity degree of participant.RE _pabe worth higher, rank price is just relatively lower.Rank price S _ionly for rank, participant's real income is still b _i.Platform is according to rank price S _irank, selects m name participant in order successively, and meets

Step 5: platform notice competitive bidding successful m name participant, m name participant sends to platform by after information.As long as transmission data, platform will give the virtual award of participant (that is: badge).

Step 6: the confidence level of platform to data calculates.Specifically comprise:

1. the quality coefficient Q of data is calculated by existing video image quality detection technique, sound quality detection technique etc.Wherein the value of Q belongs to 0 ~ 1.

2. calculate the confidence level T of participant. wherein TE is the set of the speciality of required by task, and PE is the set of the speciality that participant has, and n gathers the number comprised.

3. calculate the confidence level Tod (trustworthynessofdata) of data, Tod=a × Q+ (1-a) × T.Because weight is higher shared by the quality of data, so get a=0.7 here.The value of Tod belongs to 0 ~ 1.

Step 7: publisher receives the data that user sends, carries out evaluation e to data _pu.

Wherein

Here as the e of publisher _puvalue belong to favorable comment or in when commenting, proceed to step 8:

If e _puduring for poor commenting, then enter Game Authentication mechanism, comprising:

I. platform carries out the authentication method based on game.Namely platform is marked to data by recruiting player, gives the certain amount of money of player and rewards, the scoring of player and the scoring of publisher are compared, determine whether publisher has the behavior of malice evaluation.

Ii. n name player recruited by platform, is c (that is: c is constant) to the amount of money award of every player.The wherein selection demand fulfillment G=n × c≤0.5b of n _i, to alleviate the extra amount burden that participant bears.

Iii. player marks to data, draws wherein e _gvalue belong to 0 ~ 1.

Iv. platform is by e _gwith e _puvalue compares, and judges whether publisher is malice scoring.

V., after thinking after platform is passed judgment on that publisher is correct scoring, platform will return publisher b _ithe amount of money of+nc.

After thinking after platform is passed judgment on that publisher is malice scoring, then the expense of publisher is b _i+ 2nc, wherein b _ifor paying participant m _i, nc pays player, and another nc is as the extra punishment of platform for publisher.

Step 8: when publisher give be evaluated as non-difference comment time, platform is by amount of money b _ibe given to participant m _i.

Step 9: platform upgrades the popularity degree of both sides according to the trading situation of both sides.

As shown in Figure 2, as publisher for difference is commented, when carrying out playing mechanism, the scoring e collected by player _gneed the scoring e with publisher _pucompare, judge whether publisher is malice scoring with this.Make Δ=e _g-e _pu, scoring used when making e represent that actual computation popularity is spent.Draw following result, comprising:

1.e _g∈ (0,0.6], e _pu∈ (0,0.6] time, when namely player and publisher are poor commenting, illustrate that publisher is correct scoring.

As Δ=e _g-e _pu>=0.3, get e=(e _g+ e _pu)/2,

(wherein e is for calculating the scoring of using when popularity is spent, and all e are below)

As Δ=e _g-e _pu<0.3, gets e=e _pu

2.e _g∈ (0.6,0.8], e _pu∈ (0,0.6] time, when namely player's scoring is commented in being

As Δ=e _g-e _pu>=0.5, get e=e _gillustrate that publisher is malice scoring

As Δ=e _g-e _pu∈ [0.2,0.5), get e=(e _g+ e _pu)/2, now as e=(e _g+ e _pu)/2 ∈ (0.6,0.8] time, illustrate that publisher is malice scoring; As e=(e _g+ e _pu)/2 ∈ (0,0.6] time, illustrate that publisher is correct scoring.

As Δ=e _g-e _pu∈ (0,0.2), gets e=e _pu, illustrate that publisher is correct scoring.

3.e _g∈ (0.8,1.0], e _pu∈ (0,0.6] time, namely player is favorable comment, when publisher is for poor commenting,

Get e=e _g, illustrate that publisher is malice scoring.

As shown in Figure 3, the input of participant's popularity degree comprises the scoring of publisher and the confidence value Tod of data.Specifically comprise:

(1) as the evaluation e that publisher provides _pufor favorable comment or in comment time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein threTod is the judge threshold value of data reliability, gets threTod=0.5, k ₁and k ₂be the constant that determined by platform and all belong to 0 ~ 1.

(2) as the evaluation e that publisher provides _puduring for poor commenting,

A. when evaluation result for publisher for correctly evaluate time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein e is that platform is to e _gand e _pucompare the result of rear generation.

B. when evaluation result is evaluated for malice for publisher,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2}{1+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2}$

As shown in Figure 4, the input of publisher's popularity degree comprises credit grade.

First the current credit grade of this subtask publisher is defined (wherein non-malicious difference comments number to comprise the favorable comment number of publisher, in when commenting number and publisher's difference to comment but platform is judged as the number of correctly marking), x ∈ [-1,1],

The current popularity degree of publisher is calculated according to the credit grade of publisher b, c are the constant that platform specifies, desirable b=-1, c=-5.5.Then the RE after renewal is calculated with sliding window algorithm _puvalue, then the popularity degree RE of publisher after upgrading _pu=α RE _pu+ (1-α) cRE _pu, α is the constant that platform is formulated.

Claims (10)

1. in mobile mass-rent system based on an incentive mechanism implementation method for mass-sensitive, it is characterized in that, described method comprises the steps:

Step 1: reverse auction;

Step 1-1: the task division of oneself is little task that is simple, that be easy to mobile terminal realization by publisher, is published on platform, and pays certain amount of money P _puto platform;

Step 1-2: publisher has certain restriction requirement to task, this platform needs to carry out a preliminary screening to participant;

Step 1-3: carry out the participant after preliminary screening provides oneself bidding price b according to actual conditions _i;

Step 1-4: platform is according to rank price S _icarry out postsearch screening;

Step 1-5: platform notice competitive bidding successful m name participant, m name participant sends to platform by after information;

Step 1-6: the confidence level of platform to data calculates;

Step 1-7: publisher receives the data that user sends, carries out evaluation e to data _pu; If e _pufor difference is commented, then enter step 2, otherwise enter step 2-6;

Step 2: Game Authentication;

Step 2-1: platform carries out the authentication method based on game;

Step 2-2: n name player recruited by platform, is c to the amount of money award of every player;

Step 2-3: player carries out scoring to data and draws e _g;

Step 2-4: platform is by e _gwith e _puvalue compares, and judges whether publisher is malice scoring;

Step 2-5: after thinking after platform is passed judgment on that publisher is correct scoring, platform will return the certain amount of money of publisher, and can not pay participant's amount of money; After thinking after platform is passed judgment on that publisher is malice scoring, then publisher can produce extra expense;

Step 2-6: platform is by amount of money b _ibe given to participant m _i;

Step 3: popularity degree upgrades;

Step 3-1: platform is according to the scoring of publisher, and the confidence level of data, the credit grade of publisher upgrades the popularity degree of both sides.

2. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, described step 1-1 can be divided into two steps, comprising:

The first step, the task division of oneself is the little task being simply easy to mobile terminal realization by publisher, is published on platform;

Second step, publisher's payment platform amount of money, wherein publisher payment P _pucalculating comprise: the bidding price of participant, the profit of platform, carry out the cost of authenticate of playing, P _pucomputing formula be: wherein, b _ifor m name participant bidding price separately, if there is the transaction of failure, then corresponding b _ivalue strain be 0, P _platfor the profit amount of money of platform, G is the expense that platform carries out the mechanism of playing, platform profit amount of money P _platcomputing formula be:

Wherein $per=\left\{\begin{array}{cc}0& {\mathrm{RE}}_{pu}\&GreaterEqual;{\mathrm{threRE}}_{pu}\\ (1-{\mathrm{RE}}_{pu})\&times;k& {\mathrm{RE}}_{pu}<{\mathrm{threRE}}_{pu}\end{array}\right.$

, RE _pufor the popularity degree of publisher, threRE _pufor the threshold value of platform defined, k is constant, and k gets 5% in the present invention, and the computing formula of the expense G of game mechanism is: G=n × c, and meets: G≤0.5b _i, wherein n is that platform carries out the players number of recruiting of playing, and c is the award that platform gives player.

3. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, comprise rank price S in described step 1-4 _icalculating, rank price S _icomputing formula be: S _i=b _i× (1-RE _pa), wherein, RE _pafor the popularity degree of participant, RE _pabe worth higher, rank price is just relatively lower, rank price S _ionly for rank, participant's real income is still b _i, platform is according to rank price S _irank, selects m name participant in order successively, and meets

4. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, platform notice competitive bidding successful m name participant in described step 1-5, m name participant sends to platform by after information, as long as transmission data, platform will give participant virtual award, that is: badge; Comprise the calculating to data reliability in described step 1-6, comprise two aspects, that is: 1) the quality coefficient Q of data; 2) the confidence level T of participant, to the computing formula of the confidence level Tod of data is: Tod=a × Q+ (1-a) × T, Tod belongs to 0 ~ 1, and wherein a is constant, and Q is the quality coefficient of data, and T is the confidence level of participant; Shared by the quality of data, weight is higher, a desirable 0.7, and calculated the quality coefficient Q of data by existing video image quality detection technique, sound quality detection technique etc., wherein the value of Q belongs to 0 ~ 1, and the computing formula of confidence level T is: wherein TE is the set of the speciality of required by task, and PE is the set of the speciality that participant has, and n gathers the number comprised.

5. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, in described method step 1-7, comprise the scoring e of publisher to data _pu, its standards of grading are:

If publisher is evaluated as non-difference when commenting, then platform is by amount of money b _ibe given to participant m _i, platform upgrades the popularity degree of both sides; If publisher is evaluated as difference when commenting, then Game Authentication mechanism will be carried out.

6. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, in described step 2-1, platform carries out based on the authentication method of game is that platform is marked to data by recruiting player, give the certain amount of money of player to reward, the scoring of player and the scoring of publisher are compared, determines whether publisher has the behavior of malice evaluation.

7. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, in described step 2-2, n name player recruited by platform, be c to the amount of money award of every player, that is: c is constant, wherein the selection demand fulfillment G=nc≤0.5b of n _i, to alleviate the extra amount burden that participant bears; E in described step 2-3 _gcomputing formula be: wherein e _gvalue belong to 0 ~ 1, e _girepresent n-th _iplayer is to the scoring of data for name, and belongs to 0 ~ 1.

8. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, e in described step 2-4 _gwith e _puthe comparison of value, judges that whether publisher is the comparative approach of malice scoring, and to be the actual scoring of calculating participant popularity when spending be e: a.e _g∈ (0,0.6], e _pu∈ (0,0.6] time, when namely player and publisher are poor commenting, illustrate that publisher is correct scoring;

As Δ=e _g-e _pu>=0.3, get e=(e _g+ e _pu)/2,

As Δ=e _g-e _pu<0.3, gets e=e _pu

B.e _g∈ (0.6,0.8], e _pu∈ (0,0.6] time, when namely player's scoring is commented in being

As Δ=e _g-e _pu>=0.5, get e=e _gillustrate that publisher is malice scoring

As Δ=e _g-e _pu∈ [0.2,0.5), get e=(e _g+ e _pu)/2, now as e=(e _g+ e _pu)/2 ∈ (0.6,0.8] time, illustrate that publisher is malice scoring; As e=(e _g+ e _pu)/2 ∈ (0,0.6] time, illustrate that publisher is correct scoring;

As Δ=e _g-e _pu∈ (0,0.2), gets e=e _pu, illustrate that publisher is correct scoring;

C.e _g∈ (0.8,1.0], e _pu∈ (0,0.6] time, namely player is favorable comment, when publisher is for poor commenting,

Get e=e _g, illustrate that publisher is malice scoring.

9. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, in described step 2-5 after thinking after platform is passed judgment on that publisher is correct scoring, platform will return publisher b _ithe amount of money of+n × c, wherein b _ithe amount of money be originally will to participant m _ithe amount of money, n × c carries out the expense (if namely user correctly marks, just do not need for game machine makes expense) of Game Authentication mechanism as platform; After thinking after platform is passed judgment on that publisher is malice scoring, then the expense of publisher is b _i+ 2n × c, wherein b _ifor paying participant m _i, n × c pays player, and another n × c is as the extra punishment of platform for publisher.

10. in a kind of mobile mass-rent system according to claim 1 based on the incentive mechanism implementation method of mass-sensitive, it is characterized in that, to participant's popularity degree RE in described step 3-1 _pacalculating comprise the data reliability Tod of participant and the scoring of publisher, and RE _pabelong to 0 ~ 1, its computing formula for for:

(1) as the evaluation e that publisher provides _pufor favorable comment or in comment time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1+e_{pu}\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein threTod is the judge threshold value of data reliability, gets

ThreTod=0.5, k ₁and k ₂be the constant that determined by platform and all belong to 0 ~ 1;

(2) as the evaluation e that publisher provides _puduring for poor commenting,

A. when evaluation result for publisher for correctly evaluate time,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2}{1+(Tod-threTod)\&times;k_1-(1-e)\&times;{\mathrm{RE}}_{pu}\&times;k_2},$ Wherein e is that platform is to e _gand e _pucompare the result of rear generation;

B. when evaluation result is evaluated for malice for publisher,

${\mathrm{RE}}_{pa}=\frac{{\mathrm{RE}}_{pa}+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2}{1+(Tod-threTod)\&times;k_1+e\&times;(1-{\mathrm{RE}}_{pu})\&times;k_2};$

To the popularity degree RE of publisher _pucalculating comprise the credit grade of publisher, RE _pubelong to 0 ~ 1, its computing formula is:

First the current credit grade of publisher in this subtask is defined (wherein non-malicious difference comments number to comprise the favorable comment number of publisher in this subtask, in when commenting number and publisher's difference to comment but platform is judged as the number of correctly marking), x ∈ [-1,1], calculates the current popularity degree of publisher according to the credit grade of publisher b, c are the constant that platform specifies, then calculate the RE after renewal with sliding window algorithm _puvalue, then the popularity degree RE of publisher after upgrading _pu=α RE _pu+ (1-α) cRE _pu, α is the constant that platform is formulated.